248 
Fishery Bulletin 107(2) 
Table 1 (continued) 
Geographic area 
Reporting region 
n 
Mean population sample size 
Northern British Columbia 
Taku River 
5 
34(12-65) 
North Coast 
18 
117(28-242) 
Skeena River 
13 
95(12-182) 
Grenville Channel 
6 
122 (40-242) 
Central Coast 
52 
190 (13-419) 
Rivers Inlet 
8 
79 (40-153) 
Smith Inlet 
2 
363 (226-499) 
Southern British Columbia 
Johnstone Strait 
13 
134(20-409) 
South Coast 
14 
137(15-344) 
Vancouver Island east coast 
9 
227 (167-285) 
Vancouver Island west coast 
10 
133 (43-243) 
Fraser River 
23 
151 (24-427) 
Washington 
North Puget Sound 
7 
85 (50-100) 
South Puget Sound 
3 
100(100-100) 
Hood Canal 
2 
95 (88-102) 
Strait of Juan de Fuca 
2 
100(100-100) 
Coast of Washington 
4 
91 (61-106) 
maintain a balanced design, regions included in the 
analysis required two or more populations each with 
two or more years of samples available. Regions were 
distributed around the Pacific Rim and were a subset of 
the 59 geographic regions outlined in Table 1 and Figure 
1. The specific populations included from each region 
are in shown parentheses: West Kamchatka (Hairusova, 
Vorovskaya), Western Alaska (Snake, Eldorado), Yukon 
River summer run (Gisasa, Tozitna), Southeast Alaska 
(DIPAC hatchery, Disappearance), Queen Charlotte 
Islands west coast (Clapp Basin, Mace), Queen Charlotte 
Islands east coast (Lagoon, Pallant), Northern Brit- 
ish Columbia (Ensheshese, Kateen), Grenville Channel 
(Markle, Wilson), British Columbia central coast (Bull- 
ock Channel, Quaal, Salmon), Smith Inlet (Walkum, 
Nekite), Johnstone Strait (Viner Sound, Nimpkish), 
Vancouver Island east coast (Big Qualicum, Cowichan), 
and Fraser River (Inch, Stave). Estimation of variance 
components of river drainage or region differentiation, 
among populations within drainages or regions, and 
among years within populations was determined with 
Genetic Data Analysis. 
Results 
Variation within populations 
Substantial variation was observed in the number of 
alleles at the 14 microsatellite loci surveyed in the study. 
The fewest number of alleles was observed at Oke3 (26 
alleles), and the greatest number of alleles observed at 
Onelll (149 alleles) (Table 2). Lower heterozygosity was 
observed at loci with fewer than 40 alleles. The genotypic 
frequencies at each locus conformed to those expected 
under Hardy-Weinberg equilibrium (HWE). 
The number of alleles observed displayed considerable 
variation across regional groups of chum salmon. Asian 
chum salmon were considerably more diverse than those 
in North America, with Asian populations displaying 
the greatest number of alleles at all 14 loci examined 
(P=0.0001) (Table 3). With sample sizes standardized 
to 911 fish per region, Japanese chum salmon were the 
most genetically diverse set of populations examined 
with 581 alleles observed, greater than in all other re- 
gional groups of populations. The least diverse groups 
of populations were observed in the Queen Charlotte 
Islands, the Skeena River, the east coast of Vancouver 
Island, and Washington State, with an average of 414 
alleles observed in chum salmon from these regions. 
Japanese chum salmon displayed 40% more alleles and 
Russian chum salmon 35% more alleles than did chum 
salmon from the four regions of lower genetic diversity. 
The greatest difference in diversity was observed at 
locus Onelll, with the greatest number of observed 
alleles, and Asian chum salmon displayed 80% more 
alleles than did chum salmon from the four regions of 
lower genetic diversity. Even with Onelll removed from 
the analysis, Asian chum salmon were still more diverse 
than chum salmon in all regions in North America 
(P=0.0002). 
Distribution of genetic variance 
Gene diversity analysis of the 14 loci surveyed was 
used to evaluate the distribution of genetic variation 
